A theory on why microsoft makes such crappy stuff after all these years

[Chuck Robey]

Jayton Garnett wrote:
> I used to be naive too, thinking Unix was better in all respects. Then got a
> job working in a predominately Windows  environment, and most of the
> problems I encounter are third party software issues - hardly the fault of
> Microsoft eh? Yes Unix is far more interesting as you can start with a CLI
> and build on that to make what you want, however that is also it's downfall.
> The average PC user just about knows how to turn it on, write a document,
> browse the web and print.

Well, I'm not up on the latest versions of Windows, because I also am a
unix-only person.  I HAVE made money on Windows software development, because
the money was laying there in front of me, but I have no Windows at home.  You
have made a comment about the 3rd party software being to blame, and while I
usually won't contribute to an anti-MS discussion (too easy a target), I will do
that here.  A major reason that 3rd party software on Windows has a reputation
of causing crashes is not because the software is bad, it's because the design
of the OS behind it both leaves huge holes for problems, AND hides most of the
problems.  An instance of poor OS design is the fact that access to the disk is
very difficult to do in Unix (you need root permissions to get around using the
provided filesystem drivers).  MS for years (I am not sure any more, but I knmow
for years) has allowed programs, including hordes of cheap shareware items, to
write directly tot he disk, which provides landmines on that disk for other
programs to step on.

A second point is, in Unix, something like a write to memory outside your
program's allotment gets you an error and, likely, a core dump.  This is
immediate and huge incentive for software vendors to find and fix their fatal
errors.  In Windows, having those fatal errors show themselves has always been
advertised as a weakness!  So, lots of programs break the OS environment, but
the error often won't show in that prog, it shows an hour later when simething
else hits a tripwire.  So, in Unix, making the errors obvious forces folks who
write software to fix the majority of problems.  In Windows, you get a very
uncertain environment where errors aren't fixed, but remain around and about,
waiting to hit some other prog.  Isn't it lovely how marketing can portray that
as a 'good thing'?

Having less opportunity for errors, and having an environment where many errors
make themselves immediately (and loudly) known, has always been a lot higher on
the Unix list of priorities than Windows.

Does anyone who programs currently in Windows (Windows7?) know if they've fixed
the FS access, or added stuff like all of our fatal errors (see man 2 intro)?

> 
> Explaining to a computer user non-technical how/why/when to install software
> using the ports and make or even pkg_add -r and for them to grasp this
> concept is where Windows (OS/X also) is by far the 'better' OS for these
> people, which is more than 90% of the population.
> 
> Cheers,
> Jay ;)
> _______________________________________________
> freebsd-chat at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-chat
> To unsubscribe, send any mail to "freebsd-chat-unsubscribe at freebsd.org"